Beliefs and attitudes of male and female adolescents and the risk of smoking behaviorK Kasim1, A Al-Zalabani2, ES Abd El-Moneim3, S Abd El-Moneim4
1 Department of Family and Community Medicine, Taibah Medical College, Madinah, Saudi Arabia; Department of Public Health and Community Medicine, Al-Azhar University, Cairo, Egypt
2 Department of Family and Community Medicine, Taibah Medical College, Madinah, Saudi Arabia
3 Department of Pediatrics, Sohag University, Sohag, Egypt; Department of Pediatrics, Taibah Medical College, Madinah, Saudi Arabia
4 Department of Family and Community Medicine, Taibah Medical College, Madinah, Saudi Arabia; Department of Public Health and Community Medicine, Al-Azhar University, Damietta, Egypt
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0022-3859.180546
Source of Support: None, Conflict of Interest: None
Background: Adolescent smoking relates to numerous risk factors, of which beliefs and attitudes toward smoking may play a role. The study aimed to investigate the association between beliefs and attitudes and the risk of adolescent smoking. Materials and Methods: In a school-based cross-sectional study, 3,400 students were recruited from 34 intermediate and secondary schools in Madinah City, Al Madinah Region, Saudi Arabia. Data about sociodemographics, smoking-related factors, and beliefs and attitudes toward smoking were collected using a valid and reliable self-administered questionnaire. Prevalence of smoking was estimated and the studied beliefs and attitudes were compared by smoking status and sex using appropriate statistical analyses including multivariate logistic regression. Results: Of the 3,322 respondents, 33.02% (38.9% males and 26.4% females) were current smokers. Beliefs and attitudes toward smoking significantly differed between smokers and nonsmokers in the studied male and female students. The adjusted risk of smoking was significantly increased among female adolescents who believed that male smokers were more attractive [odds ratio (OR) = 2.2; 95% confidence interval (CI) = 1.6-2.9] and among male smokers who believed that female smokers are more attractive (OR = 1.7; 95% CI = 1.2-2.2). The risk was also increased among all adolescents who believed that smoking lent comfort in social gatherings. Belief that smoking is harmful, however, was negatively associated with the risk of smoking, particularly among females (OR = 0.55; 95% CI = 0.35-0.91). Conclusions: The study revealed a considerable high prevalence of smoking among male and female adolescents. Addressing the beliefs and knowledge about smoking during childhood is crucial in any antismoking program.
Keywords: Adolescents, attitude, beliefs, Saudi Arabia, smoking
Tobacco use and its health consequences is one of the most serious public health problems worldwide.  Evidence is accumulating that smoking increases the risk of cancers and cardiovascular and respiratory diseases.  Despite proven negative health effects of smoking, it is becoming more prevalent, particularly among adolescents, in the last few decades.  Most researches concerned with adult smoking have reported that the majority of smokers begin to smoke early in the adolescent period, before the age of 18 years. ,, Adolescence is a critical period characterized by psychological and behavioral changes that may affect adolescents' smoking behavior. Several beliefs and attitude are shaped as well in this period.  This makes school years a crucial period to study not only the smoking prevalence and predictors but also the beliefs and attitudes of adolescents toward smoking during this period.
Several recent studies have addressed the prevalence and predictors of smoking among Saudi adolescents. ,,,,, The highest reported prevalence rates of smoking in Saudi adolescents were 37% in Jeddah,  and 31% and 29% in Riyadh, Riyadh Province, Saudi Arabia. , The most important predictors associated with the risk of adolescent smoking revealed in those studies were parental and friends' smoking behavior, teachers' smoking, academic achievement, school level, age, and sex of adolescents. With the exception of one study,  most published reports on the Saudi population have failed to clearly address the role of the beliefs and attitudes toward smoking as risk factors for adolescent smoking behavior. Investigating and knowing the adolescent beliefs and attitudes about smoking may aid in the reduction of smoking among them and in the design of appropriate and effective smoking prevention and control programs targeting this important segment of the population.
The aim of this study was to identify the beliefs and attitudes toward smoking among smoker and nonsmoker male and female adolescents, and to investigate the association between belief- and attitude-related factors and the risk of adolescent smoking in Madinah City, Al Madinah Region, Saudi Arabia.
This school-based, cross-sectional study analyzed data from 3,322 students from intermediate and secondary schools in Madinah City, Al Madinah Region, Saudi Arabia during the 1st semester (September to January) of the academic year 2013/2014. A multistage, stratified cluster sampling procedure was employed, in which schools in Madinah City, Al Madinah Region, Saudi Arabia were defined in strata according to their level (intermediate vs secondary), status (public vs private), and students' sex (male vs female) with the final sample proportional to the size of the stratum. Within each stratum, a cluster sampling was implemented in which the primary sampling unit was the school. Finally, within each selected school, one class from each grade was randomly selected. All students in each selected class were invited to participate.
The primary calculated sample of this study was 780 students based on the average of the estimated smoking prevalence among school students in previous Saudi studies (20-30%), and a precision of 0.03 and confidence interval (CI) of 95% was assumed. To obtain the same level of accuracy in both male and female students as well as in intermediate and secondary schools, the sample size was quadrupled to be 3,120 students. Bias of nonresponse and missing data were also taken into consideration to arrive at the sample size of 3,400 students.
The study data were collected through valid, structured, self-administered, anonymous questionnaires. The used questionnaire was based on the Global Youth Tobacco Survey (GYTS) questionnaire.  The GYTS is a school-based instrument consisting of 56 core questions designed and validated to gather data on the prevalence of cigarette smoking and its associated risk factors. The questionnaire addresses questions about the smoking status, smoking-related factors, sociodemographic factors, beliefs and attitudes toward smoking, and behavioral characteristics of the respondents.
Arabic translation of the questionnaire was done and verified by back translation performed by a different bilingual person who had not seen the original English language version. A pilot testing of the questionnaire was performed on a group of students by asking them to fill in the questionnaire and give feedback about their understanding of the questions and any ambiguity. Public health and tobacco-control experts reviewed the results of the pilot testing as well as the back translation. The construct and content validity of the used Arabic questionnaire was obtained from discussions with them and it exceeded 98%.
The study questionnaire was explained to the students by trained public health personnel and then the students were asked to fill the questionnaire anonymously. Participation in the study was voluntary and the school officials were clearly informed about the aim and scope of the study. The study was announced to parents, giving them the chance to refuse the participation of their children by informing the school officials. Before the administration of the study questionnaire, the interviewer read out the consent form in order to obtain written consent. The confidentiality and privacy of the collected data were insured throughout the study. Finally, the study protocol was approved by the Deanship of Scientific Research Ethics Committee at Taibah University, Madinah City, Al Madinah Region, Saudi Arabia.
Smoking status, the dependent variable, was assessed in the study questionnaire by the following questions: "Have you ever tried smoking a cigarette, even once?," "During the past 30 days, how many days did you smoke cigarettes?" followed by "During the past 30 days, on the days you smoke, how many cigarettes did you usually smoke?" and "How old were you when you first tried a cigarette?" Smokers were defined as students who had smoked at least once in their lifetime while never smokers were those who had never tried smoking in their lifetime. The current smokers were defined as students who had smoked at least once in the past 30 days. 
The main independent variable in this study was the factors related to beliefs and attitudes toward smoking. These factors were assessed by asking the student to select one of the three response options concerning the studied belief. The adolescent's belief about the harmful effects of smoking was assessed by the question "Do you think that smoking harms your health"--"no," "do not know," and "yes." Responses of "no" and "do not know" were grouped as "no" and categorized as the reference group. Analogous procedures were applied for all studied belief and attitude factors.
Other explanatory variables included in this study were: i) Sociodemographic characteristics - age in years (≤13, 14, 15, ≥16), sex (male vs female), school level (intermediate vs secondary), school status (public vs private), pocket money (≤SAR300 vs >SAR300), parent education (illiterate, basic education, and university or higher education); ii) smoking-related factors; parental smoking (none, one parent smokes, both parents smoke), best friends smoking (none, some, most, or all), and having seen tobacco adverts in mass media (yes vs no).
The data were analyzed using the Statistical Analysis System software package.  Descriptive statistics were used to estimate smoking prevalence, and chi-square test was used to compare adolescents' beliefs and attitudes toward smoking by their smoking status and sex. P values ≤0.05 were considered as evidence of statistical significance. Multivariate logistic regression analyses were used to investigate the association between adolescent smoking and the studied beliefs and attitude factors while controlling the possible confounders. To avoid the confounding effects of factors known to be associated with adolescent smoking, the following potential confounding factors were considered: age, sex, school level and status, pocket money, friends and parental smoking, parental educational level, and family composition.
A total of 3,322 students were included in the study analyses. The response rate was 97.7%, with no significant school level and sex differences. The prevalence of smoking among the studied adolescents was 33.02% (1,097/3,322), and the majority of smokers in this sample (84%) reported to have started smoking before the age of 14 years with no significant sex difference. [Table 1] presents the prevalence of smoking in the studied adolescents by their characteristics. The prevalence of smoking was significantly high among adolescents who lived with neither parent (57%), those reported that their best friends smoked (53%) and those who reported that one or both parents smoked (45%). Concerning adolescents' sociodemographic characteristics, a significantly high prevalence was also detected among adolescents aged ≥16 years, reporting a monthly pocket money >SAR300, males, secondary and private school students, and those in the illiterate parents group.
[Table 2] showed adolescents' beliefs and attitudes toward smoking by their smoking status and sex. There were statistically significant differences between smoker and nonsmoker adolescents regarding most of the studied beliefs and attitudes about smoking. The percentage of adolescents who believed in the harmful effects of smoking was significantly higher among nonsmokers compared to smokers, particularly among females (95% vs 89%, P < .0001). The percentages of adolescents who believed that smokers have more friends, were more attractive, and that smoking helps people feel comfortable in social gatherings were statistically and significantly higher among smokers compared to nonsmokers. These differences also remained statistically significant when the analysis was stratified by sex, with the exception of the variable "belief that boy smokers have more friends." The stratified analyses by sex revealed that nonsmoker female students had a lower approval rate about male and female smokers, compared to nonsmoker male students (9% vs 15% in case of female smokers' approval and 17% vs 19% in case of male smokers' approval).
[Table 3] presents the risk of adolescent smoking associated with beliefs and attitude factors considering adolescents' sex. A significantly lower risk of smoking was detected among adolescents who believed in the harmful effects of smoking and it was more marked among females [odds ratio (OR) = 0.55%; 95% CI = 0.35-0.91] when compared to those who did not believe in the harmful effects of smoking. Compared to adolescents who did not believe that smokers look more attractive, the risk of smoking was significantly increased among adolescents who believed that smokers were more attractive, with the highest risk found among females who believed that male smokers were attractive (OR = 2.10; 95% CI = 1.65-2.91), and among males who believed that females smokers were attractive (OR = 1.65; 95% CI = 1.17-2.10). A significant positive risk was also detected among adolescents who believed that smoking helps people feel comfortable in social gatherings when compared to those who did not believe in it with an adjusted OR of 2.70 (95% CI = 2.10-3.27) among all studied adolescents, 2.95 (95% CI = 2.22-3.87) among male adolescents and 2.50 (95% CI = 1.90-3.40) among female adolescents.
Our results have shown that about a third of intermediate and secondary school adolescents in Madinah, Al Madinah Region, Saudi Arabia smoked. Although the proportion who smoked was higher among male and secondary school students, it also remained alarmingly high among female and intermediate school students. Further analyses showed that in both secondary and intermediate schools, male smoking outweighed female smoking. A similarly high prevalence of adolescent smoking was also reported in previous Saudi studies. ,, The reported overall prevalence rates of adolescents smoking in those studies ranged 29-37% compared to 33.02% in this study. Smoking prevalence with the same patterns of sex and level of education in adolescents was observed in a school-based study conducted in Riyadh, Riyadh Province, Saudi Arabia. 
Based on the theory of planned behavior (TPB), behavior is influenced by intention, attitude, subjective norm, and perceived behavioral control. The latter factor represents one's beliefs about the advantages/ease and disadvantages/difficulties coupled with adopting a behavior.  TPB has been widely used in designing theory-based intervention programs to reduce smoking and it was shown that perceived behavioral control exerts a powerful influence both on the intention and the behavior of smoking in a way that is presumed to be culture-specific.  Although several local studies ,, have addressed a number of social and psychological risk factors associated with adolescent smoking, they have failed to address the role of adolescents' beliefs and attitudes toward smoking as risk factors for adolescent smoking in this country. This study aims at filling that gap.
The belief that smoking has harmful health effects has significantly reduced the risk for adolescent smoking by more than a third. On stratified analysis, and although lower risk of smoking was demonstrated in both male and female adolescents who believed in the harmful health effects of smoking, it reached statistical significance only in female adolescents. A similar lower risk was also reported in previous studies from different cultures, ,,, where the perception of hazards and long-term consequences of smoking were negatively associated with adolescent smoking. These findings reflect the importance of disseminating the information about the health hazards of smoking and discouraging tobacco use among adolescents.
The adolescents' perception that smoking enhances social image and physical image could influence the initiation and maintenance of smoking. Our findings revealed that an overwhelming majority of the studied males as well as females, whether smoking or not, believed that smoking affected the body weight. In a previous clinical trial, students submitted to smoking cessation programs were concerned with weight gain and body image following smoking cessation.  Other previous reports have also revealed a positive association between weight concerns and smoking among female adolescents but not among male adolescents. , However, our findings suggested that the belief that smoking affects body weight increased the likelihood of smoking among male adolescents more than it did in female adolescents.
While the risk of smoking behavior in female adolescents was not associated with the belief that smokers, boys or girls, have more friends, in male adolescents, unexpectedly, the risk of smoking behavior showed a significant association with the belief that girl smokers and not boy smokers have more friends. This might indirectly reflect that male adolescent smokers expect forming a friendship with a female adolescent to be easier if she smokes. A previous study  reported no significant difference between male and female adolescents regarding their belief that smokers have more boyfriends or girlfriends. In that study, the authors have not stratified the results by smoking status as this study did. However, the ability to form friendships substantially depends on the social and cultural milieu that adolescents are living in. A significant positive association was found between adolescent smoking and the belief that smoking individuals were more attractive. The risk for developing a smoking habit was high among adolescent females who believed that male smokers were more attractive and among adolescent males who believed that female smokers were more attractive. Nonsmoker female students reported a lower approval rate about male or female smoking, compared to nonsmoker male students. These rates about adolescents' attitude toward male and female smokers were consistent with those reported in the previous African  and Asian  studies. A positive image of male smokers appeared to fit into the social, traditional, and cultural concepts in these communities such that smoking may be viewed as an acceptable male social behavior while being considered unusual for females. The previous findings point to two important implications on antismoking programs directed to adolescents. First, antismoking messages should be designed differently for male and female adolescents. Second, addressing the image of the opposite sex in such programs seems to be important.
Male and female adolescent smokers alike believed more than their nonsmoker counterparts that smoking helped people feel more comfortable in social gatherings. The study findings also revealed a significant positive risk of smoking among adolescents who believed that smoking helps people to feel comfortable in social gatherings. This risk was most pronounced in males. This was concomitant with the findings from other international and local studies, in which the majority of adolescent smokers reported smoking as a mean of comfort in social gathering , that was in turn reflected as a main reason for smoking among adolescents. 
Although the previous studies ,, have not stratified results by sex, their findings together with the result of this study support the argument that adolescents, particularly males, smoke for pleasure and to feel comfortable at celebrations, parties, and other social gatherings. 
The present study has a number of strengths that include being school-based with a relative large sample size, high response rate and precisely estimated risks as indicated by the observed narrowness of their CIs, which supports the robustness of the study findings. Furthermore, the study presented the risk of smoking behavior associated with the studied variables, separately, in the studied male and female students. Limitations of this study should also not be overlooked. The data collection in the study was based on self-completion of the GYTS questionnaire. The validation of self-report via biochemical tests was not feasible due to logistical and cultural constraints. This could have underestimated the actual prevalence of smoking in the studied adolescents. However, the previous studies , have reported a high reliability of results on teenager smoking when self-reported questionnaires were administered anonymously. Also, the causality between the studied factors and adolescent smoking cannot be determined because of the cross-sectional design of the study.
In summary, this study revealed a considerably high prevalence of adolescents smoking in Madinah City, Al Madinah Region, Saudi Arabia. Believing in the health hazards of smoking was significantly associated with a lower risk of adolescent smoking while believing that smokers were more attractive, especially if they were from the opposite sex, and the belief about more comfort in social gatherings increased the likelihood of attracting adolescents to smoking. The beliefs that smoking affects the body weight and smokers have more friends seemed to affect the risk of adolescent smoking differently in males and females. Given the abovementioned limitations, the findings of this study have important implications for smoking prevention and control among adolescents. Without modifying such beliefs and attitudes about smoking among adolescents and without attention to the differences between males and females, the prevalence of smoking among this important segment of the population is not expected to decline. Since the majority of smokers in this study have reported initiating smoking before the age of 14 years, a logical suggestion for smoking prevention would be to work early on modifying such beliefs and attitudes by disseminating effective structured messages about smoking in the childhood period through professionals who work with children such as pediatricians and family physicians, and to include knowledge about the health hazards of smoking in the educational curriculum as early as primary education.
The authors would like to thank all intermediate and secondary school students who participated in this study. The authors also acknowledge the teachers and administrators of all studied schools for their assistance and cooperation. The authors are also grateful for the Deanship of Scientific Research, Taibah University, Madinah, Al Madinah Region, Saudi Arabia for the support.
Financial support and sponsorship
This project was funded by the Deanship of Scientific Research, Taibah University, Madinah, Saudi Arabia, under grant no. 1758/1433.
Conflicts of interest
The authors declare that there have no competing interests regarding the publication of this manuscript.
[Table 1], [Table 2], [Table 3]